The EAI-2300 Hailo-8™ MXM AI Acceleration Module brings powerful, efficient AI processing to space- and power-constrained edge systems. Featuring dual Hailo-8 NPUs, the module delivers 52 TOPS of INT8 performance while consuming up to 20W. Designed for industrial environments, it supports wide operating temperatures and integrates seamlessly with Advantech’s Edge AI SDK, Hailo’s Dataflow Compiler, and ready-to-use TAPPAS examples—accelerating model conversion, deployment, and scaling across diverse AI workloads.
FEATURES:
- Dual Hailo-8 NPUs delivering up to 52 TOPS of AI acceleration.
- Slim MXM 3.1 Type A form factor for easy integration into edge systems.
- Low power consumption (max. 20W) for energy-efficient AI inference.
- Wide operating temperature range from -40°C to +70°C with optional fansink.
- PCIe Gen3 x8 interface for high-throughput data transfer.
- Broad framework support including TensorFlow, ONNX, PyTorch, and Keras.
- Integrated AI software ecosystem with Advantech Edge AI SDK, Hailo Dataflow Compiler, and TAPPAS examples.
Advantech Automation EAI-2300 Hailo-8 MXM AI Acceleration Module
Advantech Automation Technical Datasheets
